• Announcements

    • Reminder - MoDaCo position on illegal content   07/30/15

      ILLEGAL CONTENT I'd like to just reaffirm MoDaCo's position regarding piracy and illegal content in the light of some recent questions / postings. Posts will be censored by myself or my moderation team if the contain or link to: Illegal / pirated / cracked software or sites that host such softwareNintendo emulators / ROMs or sites hosting them (in light of Nintendo's legal stance)CUSTOM ROMS You may discuss and post links to custom device ROMs on MoDaCo, provided the following rules are adhered to: ROMs must not contain any illegal 3rd party software (this includes trial versions included without permission)ROMs must give full credit to the original authorISSUES If you have any issues with this policy, please contact PaulOBrien directly via PM.
    • Reminder: Selling items on the forum directly is not allowed   07/30/15

      Please note that selling items on the forum directly is not allowed by the forum rules. There is a forum for eBay auctions whereby you can list the items on eBay and link to them there. This is the ONLY forum for this type of activity. You may also advertise links to the eBay forum in your signature. Please note that selling directly in contravention of these rules will result in a warning / suspension / ban.

Advent Vega kernel source code now available!

2862 posts in this topic

Posted · Report post

Hi Eduardo,

i bumped the GFX memory back up to 128 and i still have the same crash. is it possible that these libs want even more then 128MB ram for GPU? the other libs worked just fine at 128 MB.

i also checked all the libs and none of the nvomx libs are missing any symbols either.

0

Share this post


Link to post
Share on other sites

Posted · Report post

Hi Eduardo,

i bumped the GFX memory back up to 128 and i still have the same crash. is it possible that these libs want even more then 128MB ram for GPU? the other libs worked just fine at 128 MB.

i also checked all the libs and none of the nvomx libs are missing any symbols either.

It's the AVP kernel module that is failing to allocate memory. Check your carveout memory assignments. Perhaps, the problem is not the amount of memory set aside to graphics, rather than, the amount of memory set aside to AVP... You can use the p10an01 mem mappings, that we know that work...

0

Share this post


Link to post
Share on other sites

Posted · Report post

Eduardo,

Is there an easy way to test and support more huawei modems? I still have to test my e153 USB modem with the latest ril. Last time I tested it, it did not work. I guess it should be something like changing/adding at strings somewhere. Or am I mistaken?

In fact, i do suspect that it would be easier than that. Most Huawei modem use exactly the same AT commands. The main difference is the amount of USB serial port emulated, and some modems instead of using a ppp session over ttyUSB0, they just create a new network interface. The RIL should support those.

But, something missing is autodetection of ports.

The E169 3G modem, when plugged in, creates 3 virtual serial ports

ttyUSB0, used for modem data and ppp links

ttyUSB1, used as a diagnostics port (not used by RIL right now)

ttyUSB2, used to send commands to the modem, even if a ppp session is running. But, you can alos use ttyUSB0 to send commands when no ppp session is running. ppp sessions are used for 3G connectivity

The EM770W 3G modem, , when plugged in, creates 5 virtual serial ports:

ttyUSB0, used for modem data and ppp links

ttyUSB1, used as a diagnostics port (used by RIL to support voice, as there is an AT command to redirect audio through this port)

ttyUSB2, used to send commands to the modem, even if a ppp session is running. But, you can alos use ttyUSB0 to send commands when no ppp session is running. ppp sessions are used for 3G connectivity

ttyUSB3: Used to get NMEA GPS data (the EM770W also has a GPS function builtin)

ttyUSB4: Used to send propietary AT control commands to the GPS

There are some other Huawei modems that are not using this port layout. And some others, besides creating ttyUSB ports, also create a network interface, and use it instead of the ppp link. What you must find out if the number of ports your modem is creating. The first thing we need to know is the port used to control the modem. And we must also know if it requires a ppp link or a network interface to create a 3G connection. All those thing can easily be checked by plugging the modem and looking at dmesg for the ports being created...

0

Share this post


Link to post
Share on other sites

Posted · Report post

Ideally, if we want to support external 3G modems, we need usb_modeswitch running in Android. The problem is that most 3G modems (Huawei) have this 0-setup /cd emulation mode, and when you plug it into the tablet, they are detected as pendrives. We need to switch them to modem mode. After that, the ril could read the configuration and try to properly setup ports...

0

Share this post


Link to post
Share on other sites

Posted · Report post

Here is the compiled RIL

Thank you, Scanno - I downloaded and installed, just finished the tests.

BTW: To quickly change RILs I did the following: mkdir /data/local/ril-test && mv /system/lib/original-libhuaweigeneric-ril.so /data/local/ril-test && ln -s /data/local/ril-test/libhuaweigeneric-ril.so /system/lib/libhuaweigeneric-ril.so , ln -s /data/local/ril-test/original-libhuaweigeneric-ril.so /data/local/ril-test/libhuaweigeneric-ril.so , so that I just need to change that last link, if I want to switch RILs. Is there a better or preferred method?

The only thing missing would be to fix the sms issue that we are getting delayed by one sms... I have to figure out what it is happening... If you erase all the sim messages at least once, it starts working reliably...

I tested this by deleting all messages from a SIM card, but the off-by-one problem persisted. I then used a brand new SIM and everything work perfect. I then shut down the tablet, removed and reinserted the SIM, started the tablet: Off-by-one error was back.

On a side note, your algorithm to detect swapped audio nibbles now works close to perfect, the tablet is now a really usable phone.

0

Share this post


Link to post
Share on other sites

Posted · Report post

Thank you, Scanno - I downloaded and installed, just finished the tests.

On a side note, your algorithm to detect swapped audio nibbles now works close to perfect, the tablet is now a really usable phone.

I love the picture of you holding the Vega to your ear, as you walk along the street! (I know you do not, but you may have picked up that I like my little jokes).

0

Share this post


Link to post
Share on other sites

Posted · Report post

I love the picture of you holding the Vega to your ear, as you walk along the street!

There is even a technical term for that: "Talking to a pizza box"

0

Share this post


Link to post
Share on other sites

Posted · Report post

Thank you, Scanno - I downloaded and installed, just finished the tests.

BTW: To quickly change RILs I did the following: mkdir /data/local/ril-test && mv /system/lib/original-libhuaweigeneric-ril.so /data/local/ril-test && ln -s /data/local/ril-test/libhuaweigeneric-ril.so /system/lib/libhuaweigeneric-ril.so , ln -s /data/local/ril-test/original-libhuaweigeneric-ril.so /data/local/ril-test/libhuaweigeneric-ril.so , so that I just need to change that last link, if I want to switch RILs. Is there a better or preferred method?

I tested this by deleting all messages from a SIM card, but the off-by-one problem persisted. I then used a brand new SIM and everything work perfect. I then shut down the tablet, removed and reinserted the SIM, started the tablet: Off-by-one error was back.

On a side note, your algorithm to detect swapped audio nibbles now works close to perfect, the tablet is now a really usable phone.

Seems android is not acking the last received message. A logcat -b radio of the process of receiving a message would be great here :)

0

Share this post


Link to post
Share on other sites

Posted (edited) · Report post

Seems android is not acking the last received message. A logcat -b radio of the process of receiving a message would be great here

Coming up, but will take a few hours. It's saturday midnight here, and the significant other half doesn't like me fooling around with the Tablet. :P

Edited by LeoP
0

Share this post


Link to post
Share on other sites

Posted (edited) · Report post

Seems android is not acking the last received message. A logcat -b radio of the process of receiving a message would be great here

Logcat is at http://www.mediafire...k15v9q53l7cu37y

Edited by LeoP
0

Share this post


Link to post
Share on other sites

Posted · Report post

Excellent!... BTW, i have exactly the same setup, i added the em770W module to a pov mobii. But, your baseband version is newer... But having exactly the same issues as mine...

I will take a look at the logcat radio :)

0

Share this post


Link to post
Share on other sites

Posted · Report post

Well, the problem with the Sms was that the huawei modem reports the memory location (register( in the SIM where the recived message was stored, but, it reports it as a 0 based index (0=first location, 1=second, etc). It also expects to receive the same index (0 based) to delete sms in Sim. And also, expects to receive the same index (0 based) to read sms from Sim using the AT+CMGR command.

But, Android RIL does not read messages from Sim using CMGR. Instead, it directly accesses the Sim card using Sim access commands (AT+CRSM). Those commands expect a 1-based register number... So, when the first message arrived, The huawei modem told Android a new message was received and was stored in index 0. Android tried to read it using the AT+CRSM, asking for register 0, which is an invalid register. So, the huawei modem refused to return the message text. Android aborted the message read procedure, leaving the message in a unread state

When the second message arrives, the huawei modem stores it in the 2nd register, and tells android it was received and stored at index 1 (command (+CMTI:) , then android read the record number 1, using AT+CRSM, so it ended reading the previous one message, and marking it as read.

The same thing happended with the 3rd message.

What i did was to add 1 to the index the huawei modem sends (+CMTI), so from now on, it will be 1-based, matching what the AT+CRSM command expects.

Attached the source code of the modified ril, the compiled version of it, and also, the cleaned up radio log with just the important things to understand the problem

huawei-generic-23set.rar

libhuaweigeneric-ril.rar

radiolog.txt

0

Share this post


Link to post
Share on other sites

Posted · Report post

What i did was to add 1 to the index the huawei modem sends (+CMTI), so from now on, it will be 1-based, matching what the AT+CRSM command expects.

ejtagle, you are a genius! I downloaded, installed tried it out: Works absolutely perfect!

You, sir, are amazing!

1

Share this post


Link to post
Share on other sites

Posted · Report post

Using a bluetooth headset with mic, this tablet could be used as a complete phone .. ;)

0

Share this post


Link to post
Share on other sites

Posted (edited) · Report post

Using a bluetooth headset with mic, this tablet could be used as a complete phone ..

No other choice right ? Wired headset? builtin mic/speakers? ?

Edited by Daedric1383
0

Share this post


Link to post
Share on other sites

Posted · Report post

No other choice right ? Wired headset? builtin mic/speakers? ?

You could use a wired headphone (external mic is not supported by the Advent hardware... The only other choice would be an usb headset. Supposedly, it should be supported by android) ... Built in mic/speakers are also supported. I would not call audio quality the best, but certainly it can be used. Some people use the same SIM on the tablet and on their phones. Its a plus to be able to answer calls, even if only used to say they will call later , thru the real phone ;)

0

Share this post


Link to post
Share on other sites

Posted · Report post

You could use a wired headphone (external mic is not supported by the Advent hardware... The only other choice would be an usb headset. Supposedly, it should be supported by android) ... Built in mic/speakers are also supported. I would not call audio quality the best, but certainly it can be used.

Tried it out with cheapo wired earphones and internal mic, definitly works good enough. I use BT earphones and internal mic all the time for SIP and Skype, works quite well, so I suspect will work OK for phone. I use tablet a lot for work, and it's just much easier to use one device, than to use 2.

0

Share this post


Link to post
Share on other sites

Posted (edited) · Report post

I'm having a strange ocurrence, tablet seems to die in deep sleep. It just won't turn on.

I have to force turn off (hold power button) and then turn on.

Anyone has any idea? Perhaps i'm having badblocks...

UPD: I have bad blocks, in the /data partition (NAND). Somewhere between the 0 and the 32Mb.

So, i have two choices, increase /system partition so that the badblock is on /system empty space,running the risk of more yet not found bad blocks, or find a way to mark all badblocks before the first boot.

My idea was to create a jumbo CWR rom filled with garbage, to force the writing of the entire partition and then, format and flash the official one. So my question is, what's the size that the junk files should have to fill the entire block/sector/cluster ? What's the block size on Vega's nand ?

Edited by Daedric1383
0

Share this post


Link to post
Share on other sites

Posted · Report post

Hi guys, just installed the Jellybean Test beta 3-1 (nvFlash) in my PoV Mobii and it seems to look very nice. 3G and Wi-Fi came right up after install.

The tablet seems to feel sluggish though after install and thought I would reboot to see if it gets better, but unfortunately it gets stuck at the boot animation upon bootup. Tried also to re-flash to no avail. Any tricks to try to resolve this?

0

Share this post


Link to post
Share on other sites

Posted · Report post

Hi guys, just installed the Jellybean Test beta 3-1 (nvFlash) in my PoV Mobii and it seems to look very nice. 3G and Wi-Fi came right up after install.

The tablet seems to feel sluggish though after install and thought I would reboot to see if it gets better, but unfortunately it gets stuck at the boot animation upon bootup. Tried also to re-flash to no avail. Any tricks to try to resolve this?

You know the answer... Back to moddedstock and then, nvflash latest.. Don't forget to repartition SD card with gparted. All the other ways (CWM, minitool. ..) seem to cause trouble (=bootloops...)

0

Share this post


Link to post
Share on other sites

Posted (edited) · Report post

I was thinking... Why can't we enable faceunlock in vega ? ... We have the required gapps available, and, as we are now using the sd as primary storage, this should be possible (in fact, it is possible to use the SD as a complete replacement for NAND, except the boot image itself...

Gapps with facelock support... You need to modify the install script to stop validating the herring platform is used. The package includes non-neon optimized shared libs, that are a must for Tegra2 (neon extensions not supported on vega)

http://www.mediafire.com/?rkho0u1ms6dhtmh

Edited by ejtagle
0

Share this post


Link to post
Share on other sites

Posted · Report post

You know the answer... Back to moddedstock and then, nvflash latest.. Don't forget to repartition SD card with gparted. All the other ways (CWM, minitool. ..) seem to cause trouble (=bootloops...)

Repartition the sd-card? I think I'm out of the loop on this one...

0

Share this post


Link to post
Share on other sites

Posted · Report post

Repartition the sd-card? I think I'm out of the loop on this one...

Latest vegabean require3s an specially formatted sdcard, otherwise, it won't boot...

0

Share this post


Link to post
Share on other sites

Posted · Report post

You know the answer... Back to moddedstock and then, nvflash latest.. Don't forget to repartition SD card with gparted. All the other ways (CWM, minitool. ..) seem to cause trouble (=bootloops...)

Hi,Ejtagle,

I have not had any trouble partitioning my 16GB SD cards with CWM - worked each time and has not given any problems. I assume that the slower opening of apps is not a result of my partitioning using CWM.

0

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now

MoDaCo is part of the MoDaCo.network, © Paul O'Brien 2002-2015. MoDaCo uses IntelliTxt technology.